altinity-expert-clickhouse-replication

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • Prompt Injection (LOW): The skill has an indirect prompt injection surface as it processes untrusted data from the database.
  • Ingestion points: The skill reads from system.text_log (message), system.replication_queue (last_exception), and system.part_log (error).
  • Boundary markers: There are no explicit delimiters or warnings to the agent to ignore instructions embedded within the database results.
  • Capability inventory: The skill executes diagnostic SELECT queries but does not contain commands for external network access or file system modifications.
  • Sanitization: No sanitization is performed on the log messages or exception strings before they are presented to the agent for analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:43 PM